Digitizing Dance: Researchers at the Digital Worlds Institute utilize AI-driven motion analysis to benefit dancers, choreographers, therapists, and clinicians
Written by: Ryan Helterhoff (MAMC ‘23)
Associate Professor Dr. Angelos Barmpoutis and Postdoctoral Associate Wenbin Guo are leading research efforts in dance move analysis using AI-driven movement classification
“Artificial intelligence is a new revolution in how we use our digital systems,” said Barmpoutis. “It works with our existing infrastructure to automate and elevate our arts capabilities.”
With the launch of the University of Florida’s AI Initiative in 2020, the University embarked on a journey to provide researchers with unparalleled access to the world’s most powerful artificial intelligence tools. This initiative is having an impact at the UF Digital Worlds Institute, where researchers are using AI capabilities to analyze human body movement. This human movement research provides insight into a variety of fields, including the arts, health, science, and interdisciplinary fields such as dance therapy.
Digital Worlds (DW) faculty saw an opportunity to streamline the process of body movement analysis and consulted with colleagues from across the UF College of the Arts to propose an investigation of both culturally- and clinically-based human movement using AI.
Their proposal, submitted by DW Associate Professor Dr. Barmpoutis, was approved by the UF AI Research Catalyst Fund, and as a result, postdoctoral researcher Wenbin Guo joined the DW research team. The DW team was able to utilize UF’s new computing infrastructure, including access to the most powerful supercomputer in higher education, the HiPerGator. This unique supercomputer can run in-depth processes that are too complex for standard hardware, providing Digital Worlds researchers with nearly limitless computing power.
The techniques utilized in the DW movement study helped to eliminate the time-consuming watch and play process, speeding up the actual analysis while reducing the potential for human error. Scaffolding upon an existing process known as Laban movement analysis, Barmpoutis and Guo were able to develop novel methods of analyzing both new and existing imagery of human movement.
“Through AI, we can improve the abilities of those who analyze body movement,” said Barmpoutis. “This AI assistance is not intended to replace these individuals, such as a physical therapist or choreographer, but rather to augment and increase their efficiency.”
Laban movement analysis is not a new concept; it is based on the work of Rudolf von Laban, one of the founding fathers of modern dance, and had previously been expanded upon by others. It is divided into four sections: Body, Effort, Shape, and Space. While the Laban movement system provided the initial focus of this AI-driven research, the technology can be applied to other movement analysis systems as well.
The work being done at Digital Worlds has already caught the attention of Qudus Onikeku, Maker in Residence at the UF Center for Arts, Migration and Entrepreneurship. Qudus immediately saw parallels between this study and his own research on the use of AI and emerging technologies in the arts. As a Nigerian dancer and artist, he used pre-recorded African dances to test the AI technology before using live performances.
Qudus saw the potential of this research in the realm of attribution, a hot topic in entertainment and performing arts industries. Unfortunately, individual dance moves can’t be copyrighted like dance choreography can, allowing others to use and profit off the dances of individual creators. The research team hopes to provide tools for artists to digitize their dances in order to provide them with a wider range of attribution options. By using AI to analyze dance moves, artists can convert their movements into data, making it easier to sort, track, and attribute their creations.
Accessibility is a key focus in the research to make this a reality. AI computing is a data-intensive process, so the team set out to create a system that only uses a single camera to increase usability. Choreographers, physical therapists, and creators will be able to access these tools using only their smartphone.
The result? ATUNDA — a web and smartphone app that analyzes dance moves in real time and exports annotated datasets. ATUNDA, as an eventual social platform, will function as an app that allows content creators to create, protect, and store their work, as well as track remixes or usage by others.
ATUNDA’s records will be kept on another platform that Digital Worlds is familiar with: the Blockchain. Marko Suvajdzic, Director of the University of Florida’s Blockchain Lab, leads a team of researchers and also educates Digital Arts and Sciences (DAS) students on blockchain technology. The Lab is supported by a grant from the Algorand Foundation, which runs a green blockchain project aimed at advancing global economic inclusion and opportunity. The Lab, which opened earlier this year, offers blockchain technology courses and conducts research while collaborating with industry partners and participating in global blockchain conferences and events. It also supports student organizations that are exploring the emerging technology, such as the Gator Blockchain Club, which is affiliated with Digital Worlds and allows students interested in blockchain to collaborate with others and discover new applications for the technology.
Suvajdzic also works with Tim Difato, Associate Director at DW, to design courses centered around blockchain technology.
“Bringing blockchain tech into the classroom is a real game changer,” said Difato. “We’re giving students the ability to learn the ins and outs of this transformative industry, and preparing them to be key players, leaders, and innovators in this dynamic space.”
While there is still much to learn about Blockchain, researchers at the Institute are already using it in sustainable ways. Amelia Winger-Bearskin, Associate Professor of AI and the Arts at the Digital Worlds Institute, connects AI and Blockchain technologies with climate justice and indigenous solutions through the AI Climate Justice Lab. Winger-Bearskin draws on her Native American heritage to create solutions inspired by indigenous cultures that propel emerging technologies forward as tools for protecting the planet.
Like Winger-Bearskin’s cross-pollinating work involving AI, Blockchain, and indigenous practices, the Digital Worlds Institute, with its various disciplines and research specialties, has served as the convergence point for emerging technology and the arts at the University of Florida for over 20 years. Institute director James Oliverio says the pursuit of digital convergence is at the heart of DW’s work.
“As rapid advances are made in traditionally distinct technological domains, interdisciplinary research often finds ways to expand points of confluence into a larger productive convergence,” said Oliverio. “There are many previous examples of this, ranging from the confluence of electronics and communications into the mobile phone, followed by the larger convergence of the camera, music and video player, and personal computer into the smartphone. At Digital Worlds, our interdisciplinary faculty continue to creatively pursue digital convergence to expand our capabilities at the intersection of the arts, communications, and interactive technologies.”